Method and system for determining whether a drug will be effective on a patient with a disease

• US 20070231921A1
• Filed: 03/31/2006
• Published: 10/04/2007
• Est. Priority Date: 03/31/2006
• Status: Active Grant

• Current Assignee: Biodesix, Inc.
• Original Assignee: Biodesix, Inc.
• Inventors: Tsypin, Maxim, Grigorieva, Julia, Roder, Heinrich
• Granted Patent: US 7,736,905 B2
• Time in Patent Office: Days
• US Class Current: 436/173
• CPC Class Codes: A61P 35/00 Antineoplastic agents
• CPC Class Codes: G06F 18/2115 by evaluating different subsets according to an optimisation criterion, e.g. class separability, forward selection or backward elimination
• CPC Class Codes: G06F 18/24147 Distances to closest patterns, e.g. nearest neighbour classification
• CPC Class Codes: G06V 20/69 Microscopic objects, e.g. biological cells or cellular parts
• CPC Class Codes: G06V 2201/03 Recognition of patterns in medical or anatomical images
• CPC Class Codes: G16B 20/00 ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
• CPC Class Codes: G16B 40/00 ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
• CPC Class Codes: G16B 40/20 Supervised data analysis
• CPC Class Codes: Y02A 90/10 Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
• CPC Class Codes: Y10T 436/24 Nuclear magnetic resonance, electron spin resonance or other spin effects or mass spectrometry

Abstract

A process of determining whether a patient with a disease or disorder will be responsive to a drug, used to treat the disease or disorder, including obtaining a test spectrum produced by a mass spectrometer from a serum produced from the patient. The test spectrum may be processed to determine a relation to a group of class labeled spectra produced from respective serum from other patients having the or similar clinical stage same disease or disorder and known to have responded or not responded to the drug. Based on the relation of the test spectrum to the group of class labeled spectra, a determination may be made as to whether the patient will be responsive to the drug.

Claims

1. A method of determining whether a patient will be responsive to a drug or treatment, the method comprising:

obtaining a test spectrum produced by a mass spectrometer from a serum produced from a patient having a disease;

processing the test spectrum to determine a relation to a group of class labeled spectra produced from respective serum from other patients having a same or similar clinical stage disease and known to have responded to or not responded to a drug or treatment; and

determining, based on the relation of the test spectrum to the group of class labeled spectra, whether the patient will be responsive to the drug or treatment.

2. The method according to claim 1, wherein obtaining includes obtaining from a patient having the disease of cancer and the drug is an anti-cancer drug.

3. The method according to claim 2, wherein obtaining the test spectrum is from a patient who has non-small cell lung cancer.

4. The method according to claim 2, further comprising labeling the group of class labeled spectra prior to processing the test spectrum based on a known clinical benefit the anti-cancer drug had during treatment of the cancer for the respective patient.

5. The method according to claim 2, wherein the processing includes selecting at least eight peaks of the test spectrum to determine the relation of the test spectrum and group of class labeled spectra to enable determining whether the patient will be responsive to the anti-cancer drug.

6. The method according to claim 1, wherein obtaining the test spectrum is obtained from a matrix-assisted laser desorption/ionization (MALDI) mass spectrometer.

7. The method according to claim 1, wherein determining includes performing a classification algorithm on the test spectrum to determine the relation to the group of class labeled spectra.

8. The method according to claim 7, wherein performing the classification algorithm includes performing a probabilistic k-nearest neighbor computation.

9. The method according to claim 7, further comprising outputting a class label indicative as to whether the patient (i) will be responsive to the drug or treatment, (ii) will not be responsive to the drug or treatment, or (iii) no determination can be made of the responsiveness to the patient to respond to the drug or treatment.

10. The method according to claim 1, further comprising preprocessing the test spectrum prior to processing to prepare the test spectrum in accordance with processing performed on the group of class labeled spectra.

11. The method according to claim 10, wherein preprocessing includes reducing background contained in the test spectrum.

12. The method according to claim 11, wherein preprocessing further includes normalizing the background reduced test spectrum.

13. The method according to claim 12, wherein preprocessing further includes selecting peaks of the normalized, background reduced test spectrum.

14. The method according to claim 13, wherein preprocessing further includes spectrally aligning the selected peaks of the normalized, background reduced test spectrum.

15. The method according to claim 1, further comprising setting a parameter utilized in determining whether the patient will be responsive to the drug or treatment to establish a level of confidence in the form of a percentage that the patient will be determined to be responsive to the drug or treatment.

16. The method according to claim 15, wherein setting a parameter includes setting a log-rank p-value.

17. The method according to claim 1, wherein determining includes determining whether the patient will be responsive to the drug gefitinib.

18. The method according to claim 1, wherein processing the test spectrum includes selecting a plurality of differentiating peaks of the test spectrum to be processed in relation to peaks of the group of class labeled spectra.

19. The method according to claim 1, further comprising processing the group of class labeled spectra to produce a group of spectra that is substantially clinic independent and substantially mass spectrometer independent to enable the processing to be performed with a group of class labeled spectra from any clinic.

20. A method of doing business wherein the serum samples of claim 1 are used to create the test spectrum and the determination based on the relation of the test spectrum to the group of class labeled spectra is made remotely over the Internet.

21. A system for determining whether a patient will be responsive to a drug or treatment, the system comprising:

a storage device configured to store a test spectrum produced by a mass spectrometer from a serum produced from a patient having a disease and a group of class labeled spectra produced from respective serum from other patients having a same or similar clinical stage disease and known to have responded or not responded to a drug or treatment; and

a processor in communication with the storage device, the processor executing software to;

obtain a test spectrum produced by a mass spectrometer from a serum produced from a patient having a disease;

process the test spectrum to determine a relation to a group of class labeled spectra produced from respective serum from other patients having the same or similar disease and known to have responded or not responded to a drug or treatment; and

determine, based on the relation of the test spectrum to the group of class labeled spectra, whether the patient will be responsive to the drug or treatment.

22. The system according to claim 21, wherein the patients have the disease of cancer and the drug is an anti-cancer drug.

23. The system according to claim 22, wherein the cancer patients have non-small cell lung cancer.

24. The system according to claim 22, further comprising labeling the group of class labeled spectra prior to processing the test spectrum based on a known clinical benefit the anti-cancer drug had during treatment of the cancer for the respective cancer patient.

25. The system according to claim 22, wherein the processor processes the test spectrum by selecting at least eight peaks of the test spectrum to determine the relation of the test spectrum and group of class labeled spectra to enable determining whether the cancer patient will be responsive to the anti-cancer drug.

26. The system according to claim 21, wherein the mass spectrometer is a matrix-assisted laser desorption/ionization (MALDI) mass spectrometer.

27. The system according to claim 21, wherein the processor determines whether the patient will be responsive to the drug by executing a classification algorithm on the test spectrum to determine the relation to the group of class labeled spectra.

28. The system according to claim 27, wherein the classification algorithm includes a probabilistic k-nearest neighbor computation.

29. The system according to claim 27, wherein the processor further outputs a class label indicative as to whether the patient (i) will be responsive to the drug or treatment, (ii) will not be responsive to the drug or treatment, or (iii) no determination can be made of the responsiveness to the patient to respond to the drug or treatment.

30. The system according to claim 21, wherein the processor further preprocesses the test spectrum prior to processing to prepare the test spectrum in accordance with processing performed on the group of class labeled spectra.

31. The system according to claim 30, wherein the processor preprocesses the test spectrum by reducing background contained in the test spectrum.

32. The system according to claim 31, wherein the processor preprocesses the test spectrum by further normalizing the background reduced test spectrum.

33. The system according to claim 32, wherein the processor preprocesses by further selecting peaks of the normalized, background reduced test spectrum.

34. The system according to claim 33, wherein the processor preprocesses by further spectrally aligning the selected peaks of the normalized, background reduced test spectrum.

35. The system according to claim 21, wherein the processor further sets a parameter utilized in determining whether the patient will be responsive to the drug or treatment to establish a level of confidence in the form of a percentage that the patient will be determined to be responsive to the drug or treatment.

36. The system according to claim 35, wherein the processor sets the parameter by setting a log-rank p-value.

37. The system according to claim 21, wherein the processor determines whether the cancer patient will be responsive to the drug gefitinib.

38. The system according to claim 21, wherein the processor processes the test spectrum by selecting a plurality of differentiating peaks of the test spectrum to be processed in relation to peaks of the group of class labeled spectra.

39. The system according to claim 21, wherein the processor further processes a group of raw spectra to produce the group of class labeled spectra that is substantially clinic independent and substantially mass spectrometer independent to enable said processing to be performed with a group of class labeled spectra from any clinic.

40. A system of determining whether a patient will be responsive to a drug or treatment, the system comprising:

means for obtaining a test spectrum produced by a mass spectrometer from a serum produced from a patient having a disease;

means for processing the test spectrum to determine a relation to a group of class labeled spectra produced from respective serum from other patients having the same or similar clinical stage disease and known to have responded or not responded to a drug or treatment; and

means for determining, based on the relation of the test spectrum to the group of class labeled spectra, whether the patient will be responsive to the drug or treatment.

41. The system according to claim 40, wherein the patient has a cancer disease.

42. The system according to claim 41, wherein the patient has non-small cell lung cancer.

43. The system according to claim 40, further comprising means for preprocessing the test spectrum prior to processing to prepare the test spectrum in accordance with said means for processing performed on the group of class labeled spectra.

44. The system according to claim 40, further comprising means for setting a parameter utilized by said means for determining whether the patient will be responsive to the drug or treatment to establish a level of confidence in the form of a percentage that the patient will be determined to be responsive to the drug or treatment.

45. The system according to claim 40, wherein the drug is gefitinib.

46. A method of determining whether a patient will be responsive to a drug or treatment, the method comprising:

obtaining a test spectrum having a plurality of features produced by a mass spectrometer from a serum produced from a patient having a disease;

processing the features of the test spectrum to determine if a relation exists between the test spectrum and a group of class labeled spectra having differentiating peaks produced from respective serum from other patients having the same or similar clinical stage disease as the patient and known to have responded to or not responded to a drug or treatment; and

determining, based on the processed peaks of the test spectrum, whether the patient will be responsive to the drug or treatment.

47. The method according to claim 46, further comprising selecting the differentiating peaks from the group of class labeled spectra.

48. The method according to claim 47, wherein selecting differentiating peaks includes selecting at least one peak having an approximate m/z center from the list consisting of:

49. The method according to claim 48, wherein selecting differentiating peaks includes selecting at least eight of the peaks.

50. The method according to claim 48, wherein selecting differentiating peaks includes selecting twelve of the peaks.

51. The method according to claim 48, wherein selecting differentiating peaks includes selecting from the peaks having respective approximate peak widths consisting of:

57. The system according to claim 48, wherein the processor selects the differentiating peaks by selecting from the peaks having respective approximate peak widths consisting of:

52. A system for determining whether a patient will be responsive to a drug or treatment, the system comprising:

a storage device configured to store a test spectrum produced by a mass spectrometer from a serum produced from a patient having a disease and a group of class labeled spectra produced from respective serum from other patients having the same or similar clinical stage disease as the patient and known to have responded or not responded to a drug or treatment; and

a processor in communication with the storage device, the processor executing software to;

obtain a test spectrum having a plurality of features;

processing the features of the test spectrum to determine if a relation exists between the test spectrum and a group of class labeled spectra having differentiating peaks produced from respective serum from other patients having a same or similar clinical stage disease and known to have responded to or not responded to a drug or treatment; and

determine, based on the processed peaks of the test spectrum, whether the patient will be responsive to the drug or treatment.

53. The system according to claim 52, wherein the processor further selects the differentiating peaks from the group of class labeled spectra.

54. The system according to claim 53, wherein the processor further selects the differentiating peaks by selecting at least one peak having an approximate m/z center from the list consisting of:

55. The system according to claim 53, wherein the processor selects the differentiating peaks by selecting at least eight of the peaks.

56. The system according to claim 53, wherein the processor selects the differentiating peaks by selecting twelve of the peaks.